SYSTEMATIC REVIEW UPDATE Open Access

Diagnosis and treatment of Rosai-Dorfmandisease of the spine: a systematic literaturereviewPan-pan Hu, Feng Wei*, Xiao-guang Liu and Zhong-jun Liu

Abstract

Purpose: To review and summarize the clinical features, diagnosis, treatment strategies, and prognosis of spinalRosai-Dorfman disease (RDD).

Methods: RDD is also termed as sinus histiocytosis with massive lymphadenopathy. We searched the databases ofPubMed, Elsevier ScienceDirect, SpringerLink, and OVID. The keywords were Rosai-Dorfman disease and spine/centralnervous system. Research articles and case reports with accessibility to full texts regarding spinal RDD were eligiblefor the inclusion. A total of 62 articles were included, and they contained 69 cases. We extracted the information ofinterest and analyzed them using SPSS statistics package.

Results: The average age was 33.1 ± 18.3 years. The ratio of males to females was 1.9/1. Overall, 63 cases presentedwith spine-related symptoms. A total of 27 cases (39.1%) had multi-organ lesions, and 12 cases had records ofmassive lymphadenopathy. Among 47 cases who first manifested spine-related symptoms, 93.6% werepreoperatively misdiagnosed. The disease had a predilection for cervical spine (38.8%) and thoracic spine (40.3%).62.9% of lesions were dura-based. Surgery remained the mainstream treatment option (78.8%), with or withoutadjuvant therapies. Total lesion resection was achieved in 34.8% of cases. The rate of lesion recurrence/progressionwas 19.5%, which was marginally lower for total resection than for non-total resection.

Conclusion: Spinal RDD has no pathognomonic clinical and imaging features. Most cases first present with spine-relevant symptoms. Massive lymphadenopathy is not common, but a tendency for multi-organ involvement shouldbe considered. Spinal RDD has a high recurrence rate; thus, total resection is the treatment of choice. Adjuvanttherapies are indicated for multi-organ lesions and residual lesions. A wait and watch strategy is recommended forasymptomatic patients. Herein, a workflow of diagnosis and treatment of the spinal RDD is established.

Keywords: Spine, Rosai-Dorfman disease, Literature review, Clinical features, Treatment

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate ifchanges were made. The images or other third party material in this article are included in the article's Creative Commonslicence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commonslicence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtainpermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to thedata made available in this article, unless otherwise stated in a credit line to the data.

* Correspondence: puth_wf@163.comDepartment of Orthopaedics and Beijing Key Laboratory of Spinal DiseaseResearch, Peking University Third Hospital, No. 49 North Garden Rd, HaidianDistrict, Beijing 100191, China

Hu et al. Systematic Reviews (2021) 10:31 https://doi.org/10.1186/s13643-021-01581-0

BackgroundRosai-Dorfman disease (RDD) is a benign disorder, andit is also termed sinus histiocytosis with massive lymph-adenopathy [1–3]. Previous etiological studies did notarrive at any firm conclusions. Viral infections, auto-immune conditions, or gene mutations were reported tobe related to the development of RDD [3–8]. There werealso reports of family cases [9]. Symptomatically, painlessneck, axillary, and/or inguinal lymph node enlargementis present in about 80% of cases. A few patients havefever, weight loss, anemia, and malaise [1, 3]. Over 40%of cases have extranodal lesions, and the most com-monly involved organs are the skin, respiratory tract,orbit, paranasal sinuses, heart, liver and kidney, bone,and central nervous system [1–3, 9, 10]. The disease hasa tendency for recurrence and involvement of multipleorgans; Mayo clinic has referred to this condition as ahistiocytic neoplasm [3].It is estimated that 0.6–1% of RDD cases have isolated

or concurrent spinal lesions [2, 3, 9–12]. The bone, dura,and cord parenchyma are all nidus candidates. SpinalRDD has no pathognomonic clinical and imaging fea-tures. The disorder is often misdiagnosed as infection,spinal tuberculosis, extradural hematoma, meningioma,and other primary or metastatic tumors. It is hardlypossible to make the diagnosis unless histopathologicaland immunohistochemical examinations are performed.There are no consensus guidelines for the treatment ofspinal RDD. Treatment regimens are usually prescribedbased on personal experience rather than high-levelmedical evidence. Therefore, we performed a systematicliterature review specifically on spinal RDD. This studyprovides the following referential contents to peers: (a)clinical features of the disease; (b) imaging traits of RDDlesions; (c) summary of current treatment strategies andoutcomes; (d) prognosis and possible relevant clinicaland therapeutic factors.

Materials and methodsLiterature searchA literature search was performed using the followingdatabases: PubMed, Elsevier ScienceDirect, SpringerLink,and OVID. The search strategy for PubMed was pre-sented as follows: Mesh term #1: (Rosai-Dorfman)OR(sinus histiocytosis with massive lymphadenopathy): ti,ab, kw; Mesh term #2: (spine) OR (central nervoussystem). #1 AND #2. We set no limitation regarding thesearching fields and date. The language was set asEnglish. All studies retrieved were saved in the softwarepackage of EndNote X7. All retrieved literatures werecarefully screened. The inclusion criteria were (1) litera-ture regarding the cases with spinal RDD; (2) full textswere available; (3) publication types of research articlesand case reports; (4) literature with complete records of

patients’ demographic information, clinical, and imagingmanifestations; (5) references of the eligible articles werealso screened. The exclusion criteria were (1) articles innon-English languages; (2) literature with no specificcases recorded; (3) repeated literatures or same cases bysame authors in different literatures. The searching, lit-eratures screening, and selecting were performed by PHand FW independently. The data extraction was alsoperformed independently by the two authors. The resultswere checked by the other authors. All eligible articlesmet the inclusion criteria, and they did not fulfill the ex-clusion criteria (Table 1).

Risk of biasConsidering the included studies were case reports,which were ranked as level 4 evidence, an 8-item JBICritical Appraisal Checklist for Case Reports wasemployed to assess the risk of bias (ROB). If the answersfor all the 8 items were YES, this study was appraised aslow ROB; otherwise, the ROB was high. The appraisal ofthe included articles was undertaken independently byPH and WF. When disagreement occurred between thetwo appraisal results, they were then discussed and arbi-trated by the all authors.

Data collectionThe full text of the included articles was studied. Thefollowing information was extracted: (1) clinical features:genders, age, symptomatic presentations (asymptomatic,neurological deficits and/or local pain); (2) imaging fea-tures: vertebral segments (cervical, thoracic, lumbar,and/or sacrum), sites of the niduses (intrabony, dura-based, and/orintramedullary); (3) systemic involvement:lesions of lymph nodes or other organs; (4) laboratorytest results; (5) treatment regimens: wait and watch, sur-gery, and/or non-surgery therapies; (6) prognosis andtherapeutic outcomes. All the data were then transcribedonto an Excel sheet (see Supplementary Table 1). For

Table 1 The inclusion and exclusion criteria of the study

Inclusion criteria

a. Cases with spinal lesions, whether symptomatic or not

b. Full texts were availed

c. Publication types of research articles and case reports

d.Literature with complete records of patients’ demographicinformation, clinical and imaging manifestations

e. References of the eligible articles were also screened

Exclusion criteria

a. Articles in non-English languages

b. Reviews, comments, or other articles wherein no cases wererecorded

c. Duplicated literature or same cases in different articles

Hu et al. Systematic Reviews (2021) 10:31 Page 2 of 10

data not provided in original articles, we left the slotsvacant.

Statistical analysisData analyses were performed using IBM SPSS statisticsfor Windows Version 20 (IBM Corp., Armonk, NY,USA). Lilliefors test, which is an adaptation of theKolmogorov-Smirnov test, was used to examine whetherthe data were normally distributed. Data presentationwas provided in the forms of percentages, mean ± stand-ard deviation, or median (quartile range). Two-tailed un-paired Student’s t test and two-tailed Pearson’s χ2 test(or Fisher’s exact test) were utilized to make a compari-son between different groups. Statistical significance wasset at 0.05.

ResultsA total of 62 articles were included in the study, andthey contained 69 cases of spinal RDD (Fig. 1 andSupplementary Table 1) [11–72]. These articles werepublished from 1976 to 2020, and 42 articles (67.7%, 42/62) were published after 2010. Twelve included studieshad high ROB [13, 16–18, 31, 34, 46, 48, 49, 53, 56, 57],whilst 50 other studies had low ROB. The cohort

included 45 males and 24 females, at a ratio of 1.9/1.The average age was 33.1 ± 18.3 years. Follow-up infor-mation was documented in 52 cases, and the medianperiod was 12.0 (11.8) months. The median course be-fore consultation and admission was 2.1 (5.0) months(Table 2).A total of 63 cases (91.3%, 63/69) manifested spine-

related symptoms. A total of 54 cases (78.3%, 54/69)suffered from myelopathy or radiculopathy, and 31 cases(44.9%, 31/69) complained of focal pain in the lesions.Multiple organ involvement was reported in 27 cases(39.1%, 27/69), and massive lymph nodes were found in12 cases (17.4%, 12/69). Abnormal results of laboratorytests, such as elevated white blood cell count, C-reactiveprotein, and erythrocyte sedimentation rate, were de-scribed in 17 cases (24.6%, 17/69) (Table 2).The initial symptoms of 47 cases (68.1%, 47/69) were

spine-related; among them, 10 cases (21.3%, 10/47) werescreened for coexisting niduses at other organs byimaging, and 6 cases (12.8%, 6/47) had lymph node en-largement. Only 3 cases (6.4%, 3/47) were diagnosedcorrectly, while the other 93.6% of cases (44/47) weremisdiagnosed before postoperative pathological exami-nations. A total of 22 cases first presented with painless

Fig. 1 Flow diagram of the literature search, screening, and inclusion in the present study. *The other source besides the searched databases wasreferences of the included reports

Hu et al. Systematic Reviews (2021) 10:31 Page 3 of 10

lymph node enlargement or other organ-related symp-toms; among them, 16 cases (72.7%, 16/22) developedspinal lesions and symptoms afterward.Imaging work-ups revealed that the lesion manifested

T1-weighted isointensity to the cord and T2-weightedslight hyperintensity on MRI scans (Fig. 2a, b). The bonylesions were osteolytic and non-expansile and werelocated inside cancellous bones (Fig. 2c). PET/CT couldreveal multiple organ lesions, which were moderate tohigh uptake of 18-fluorodeoxyglucose (Fig. 2d).The lesion sites of spinal RDD were documented in 67

cases; among them, 26 cases (38.8%, 26/67) developedRDD lesions in the cervical spine and 27 cases (40.3%,27/67) developed RDD lesions in the thoracic spine(Table 3). Seven cases (10.4%, 7/67) had lesions in mul-tiple spinal segments. The nidus sites and lesion baseswere described in 62 cases. Intracanal dura-based lesionswere found in 39 cases (62.9%, 39/62), whereas the rates

of intramedullary and intrabony lesions were 9.7% (6/62)and 12.9% (8/62), respectively. Altogether, there were 53cases (85.5%, 53/62) with lesions focally constrained toone spinal component, namely bone, dura, or cord par-enchyma. A total of 9 cases (14.5%, 9/62) had focallyconcurrent lesions invading multiple spinal components,for example, bone and dura (Table 3).The diagnoses were made based on histopathological

and immunochemical examinations in all the includedcases, rather than on clinical and imaging manifestations.The histopathological features were histiocytic cells in a fi-brous background, with an infiltration of lymphocytes andplasma cells, and emperipolesis of surrounding cells intohistiocytic cells. The immunohistochemical phenotypewas CD68 and S100 positive and CD1a negative; thisphenotype is specific and diagnostic, according to the in-cluded reports.Treatment regimens were reported in 66 cases. A total

of 42 cases (63.6%, 42/66) underwent surgeries, 10 cases(15.2%, 10/66) received surgeries and adjuvant therapies,and 9 cases (13.6%, 9/66) received radiotherapy, steroids,and/or chemotherapy without surgeries (Fig. 3). A totalof 5 cases (7.6%, 5/66) received no specific treatment buta wait and watch strategy [17, 46, 48, 64, 69]. Amongthe 52 cases receiving surgeries, total lesion resectionwas achieved in 23 cases (44.2%), and 29 cases (55.8%)underwent decompression surgeries with or without par-tial lesion resection.A total of 54 cases reported follow-up data (Table 4).

A total of 44 cases (81.5%, 44/54) reported no recur-rence or stable residual lesions, whereas 10 cases (19.5%,10/54) experienced recurrence or progression of spinallesions. Lesions constrained to the dura, intrabonyregion, or intramedullary region were correlated withlower recurrence and progression rates than focally

Fig. 2 The imaging presentation of a case with spinal RDD. MRI scans manifested a T1 isointense (a) and T2 slightly hyperintense (b) extradurallesion on Th6. The spinous process was also involved (arrows on a and b). CT scan revealed an osteolytic lesion, but the cortex seemed intact(arrow on c). PET/CT scan was arranged. Paranasal lesions and cervical lymph node lesions were also found and the lesions manifested a highuptake of 18-FDG (red arrows on d)

Table 2 Presentation of basic data of the included spinalRosai-Dorfman casesa

Item Value

Age, mean ± standard deviation (33.1 ± 18.3) years

Course, M(QR) 2.1(5.0) months

Follow-up period, M(QR) 12.0(11.8) months

Spine-related symptoms 63(91.3%)

Focal pain at the spinal lesions 31(44.9%)

Myelopathy/radiculopathy 54(78.3%)

Lymph nodes enlargement 12(17.4%)

Multiple organ involvement 27(39.1%)

Abnormal laboratory tests 17(24.6%)

M(QR) stands for median(quartile range)aOnly data that were originally documented were included in thisstatistical analysis

Hu et al. Systematic Reviews (2021) 10:31 Page 4 of 10

unconstrained lesions (p = 0.033). Patients with systemicmultiple lesions had marginally higher recurrence andprogression rates than patients with isolated spinal le-sions (p = 0.346). Among those who underwent surgery,the rate of recurrence/progression was 17.8% and 22.2%for non-surgery cases (p = 0.754). The recurrence andprogression rates for cases after total resection and par-tial resection were 9.1% and 26.1%, respectively, with nosignificant difference (p = 0.243). For those who receivedradiotherapy, steroids, and/or chemotherapy, the recur-rence/progression rate was 28.6% and 15.0% for the pa-tients who did not receive the above therapies, with nosignificant difference (p = 0.261).Based on the review and analysis of the data regarding

on the clinical symptoms, imaging work-ups, and treat-ment and their outcomes, a preliminary algorithm wasproposed for the diagnosis and treatment of the spinalRDD (Fig. 4).

DiscussionSpinal RDD poses a great challenge for the diagnosisand treatment due to a lack of mature guidelines in clin-ical practice. According to our study, 93.6% of caseswhose initial symptoms were spine-related could not be

correctly recognized until postoperative pathologicalexamination. Computed tomography (CT)-guided biopsyresults were recorded in six cases; three cases (50.0%)failed to make a definite and correct diagnosis [17, 24,31, 34, 57, 62]. The failure rate was relatively high in theincluded reports. The possible reasons were the failureto harvest the target specimen and failure to recognize iton cellular smears. The features of cellular smear arenot diagnostic and cannot be differentiated from Langer-hans cell histiocytosis, reactive lymphadenitis, hemopha-gocytic syndrome, and malignant lymphoma [24, 57].Spinal RDD has a predilection for young and middle-

aged people, with an average age of 33.1 years and 47.8%(33/69) cases aged between 18 and 45 years. The studynoted gender variation with a predominance of males (1.9/1). Spinal RDD has no specific and pathognomonic symp-toms, although 91.3% of the included cases manifestedspine-relevant symptoms. Previous literature reported thatabout 80% RDD patients have painless massive neck or ax-illary lymphadenopathy [1, 3, 19, 21]. However, this studyfound the records of lymph node enlargement in 17.4% ofcases, which was significantly lower than that in the previ-ous literature (Table 2). It implies that spinal RDD mighthave some unique pathological features.

Table 3 Summary of lesion sites on spinal segments and lesion bases

Spinal segments (n = 67) Lesion bases/origins (n = 62)

Count Percentage Count Percentage

Cervical 26 38.80% Intramedullary 6 9.70%

Thoracic 27 40.30% Intrabony 8 12.90%

Lumbar 4 6.00% Dura-based 39 62.90%

Sacrum 3 4.50% Concurrent 9 14.50%

Multi-segment lesions 7 10.40%

Fig. 3 Treatment options for 66 cases, whose treatment regimens were recorded in the original literature. Non-surgery therapies includeradiotherapy, steroids and chemotherapy

Hu et al. Systematic Reviews (2021) 10:31 Page 5 of 10

Table 4 Correlation analysis of lesion recurrence or progression with clinical and therapeutic factors

Clinical and therapeutic factors* Recurrence and/or progression of spinal lesions

Yes No p value

Age (n = 54) 36.2 ± 21.8 34.1 ± 18.5 0.758

Focal lesion conditions (n = 51)

Focally constrained 5(11.4%) 39(88.6%) 0.033#

Focally concurrent 3(42.9%) 4(57.1%)

Involved organs (n = 33)

Single 1(6.7%) 14(93.3%) 0.346

Multiple 4(22.2%) 14(77.8%)

Treatment options (n = 54)

Surgery 8(17.8%) 37(82.2%) 0.754

Non-surgery 2(22.2%) 7(77.8%)

Surgical strategies (n = 45)

Total resection 2(9.1%) 20(90.9%) 0.243

Non-total resection 6(26.1%) 17(73.9%)

Non-surgery therapies (n = 54)

Yes 4(28.6%) 10(71.4%) 0.261

No 6(15.0%) 34(85.0%)

Non-surgery therapies include radiotherapy, steroids and/or chemotherapy. Focally constrained lesions mean isolated lesions focally constrained to the dura,intrabony region, or intramedullary region. Focally concurrent lesions mean lesions invading multiple spinal components*Only data documented in the original literature were included#Statistically significant at p < 0.05, two-tailed Pearson’s chi-squared test

Fig. 4 A preliminarily established diagnosis and treatment algorithm for spinal Rosai-Dorfman disease. ESR stands for erythrocyte sedimentationrate; CRP, C-reactive protein; CT, computed tomography; MRI, magnetic resonance imaging; PET/CT, positron emission tomography/computedtomography; TR, total resection

Hu et al. Systematic Reviews (2021) 10:31 Page 6 of 10

We found that the lesions mostly occurred in the cervicaland thoracic spine (79.1%) and that 62.9% of lesions weredura based. The imaging features of RDD are not specificand diagnostic. It is difficult to differentiate this conditionfrom spinal infections, meningioma, lymphoma and otherprimary or metastatic tumors, chronic inflammation, extra-dural hematoma, and other histiocytoses, by merely relyingon clinical and imaging manifestations.RDD has a tendency to invade multiple organs [1–3].

We noted reports of deaths due to lesions in other organs[43]. In addition, this study found that the recurrence rateof isolated spinal lesions was marginally lower than that ofmulti-organ lesions (6.7% versus 22.2%). Therefore, it isnecessary to screen all possible lesions in other organswhen dealing with spinal RDD patients. Mayo clinic re-ported a consensus on the diagnosis of systemic RDD [3].It recommended positron emission tomography/com-puted tomography (PET/CT) to rule out systemic multiplelesions in each patient. PET/CT is helpful to screenwhole-body lymph nodes and discover some uncommonand small lesions, for example, paranasal sinuses (Fig. 2d).In the included case series, only four cases had the recordsof PET/CT. This might explain the significantly lower rateof lymph node enlargement in included case series in thisstudy (Table 2). We also noticed that many included casespresented with rapidly exacerbating myelopathy, and theywere asked to undergo urgent cord decompression surger-ies, leaving no time for PET/CT. In this setting, we recom-mend imaging and ultrasonic examinations of vital organsin the intracranial and abdominal cavity.This study explored the relationship between focal

conditions of the lesions and prognosis. We found thatfocally unconstrained lesions were correlated with ahigher recurrence rate (Table 4). Generally, it is difficultto completely resect unconstrained lesions, and residuallesions have a potential for progression and recurrence.More importantly, focal involvement of multiple spinalcomponents implied that these lesions were able to infil-trate through the anatomical septa and possessed strongeraggressiveness. This finding may provide us an importantreference when dealing with spinal RDD.RDD is benign histiocytosis, and some lesions can

resolve spontaneously [47, 65]. Pulsoni et al. reviewed 40cases of untreated RDD and found 82% remittedspontaneously [73]. Rittner and colleagues reported acase with multiple whole-body osseous lesions, includingthe spine, and it was treated with non-steroidal anti-inflammation drugs [47]. After 6 months, most of the le-sions resolved spontaneously. Therefore, wait and watchis recommended for patients without symptoms andsigns of spinal instability and neurological impairment.However, for symptomatic patients, for whom cord

decompression and restoration of spinal stability areindicated, or malignancy cannot be ruled out, surgery is

the treatment of choice whenever feasible. Theoretically,total resection is a better choice, considering the ten-dency of the lesions to recur (Table 4) [3, 14, 15, 29, 30,32, 43, 44, 49, 51]. This study found that the recurrencerate of total resection was marginally lower than that ofnon-total resection (9.1% versus 26.1%). Given a longerfollow-up period (the median value of the included caseswas 12 months, see Supplementary Table 1), we inferthat the difference of the recurrence rates between thetotal and non-total resection groups would be larger andmore statistically significant. Therefore, total resection isthe recommended option, especially for those isolated,primary spinal RDD lesions. When the lesions are notcompletely resectable or systemic lesions are present, ad-juvant therapies are indicated. Adjuvant therapies mainlyinclude radiotherapy, steroids, and chemotherapy. How-ever, they were not proven beneficial in this study(Table 4). Radiotherapy has provided excellent locore-gional control for isolated lesions, but incompetent forextensive systemic involvement. Chemotherapy regimenswere usually prescribed based on personal experience ra-ther than high-level evidence, and its efficacy was notverified in the most patients [73]. The efficacy of steroidswas not supported in the case series of this study, either,especially for aged ones or patients with multiple organsinvolvement [11, 14, 31, 44, 45, 62]. Thus, surgical resec-tion, with or without adjuvant therapies, remains as aprimary modality for spinal RDD when therapeutic in-terventions are indicated.Based on a comprehensive review of the literature

on spinal RDD, we established a preliminary work-flow for the diagnosis and treatment of this disease,which may provide some assistance to our peers. Forpatients suspicious of RDD, laboratory tests of in-flammatory indexes and imaging work-ups includingCT and MRI are imperative. PET/CT is also stronglyrecommended, which can rule out systemic lesionsand, especially, the involvement of superior lymphnodes. Considering the fact that imaging modalitiesare not diagnostic, CT-guided biopsy is often neces-sary. Surgery is the treatment of choice for thepatients who present with symptoms of neurologicaldeficits and refractory local pain. Adjuvant therapies,including steroids, chemotherapy, and radiotherapy,are indicated for progressive lesions and patientswith systemic involvement. For the patients withoutneurological deficits and signs of spinal instability, await and watch strategy shall be adopted. However,this study found that the rate of lesion recurrence/progression in the spine was as high as 19.5%, imply-ing that the current treatment strategies for spinalRDD still need to be improved. Thus, high-level clin-ical research studies are required to eventually builda standard and more effective treatment guideline.

Hu et al. Systematic Reviews (2021) 10:31 Page 7 of 10

LimitationsFirstly, the included literatures in this study were mostlycase reports, which were referred to low-quality evi-dences. This can influence the quality of this study.Secondly, there may be risk of inclusion bias due tosearching strategies and the availability to some litera-ture. Thirdly, some of included literature did not havethe records of all the data of interest, which might influ-ence statistical analysis and the interpretation of theresults.

ConclusionsThe diagnosis and treatment of spinal RDD remain a hugechallenge. It has no pathognomonic clinical and imagingfeatures. Most cases initially present with spine-relatedsymptoms. Only a small number of cases have massivelymphadenopathy, fever, and other constitutional symp-toms. The lesions have a predilection for the cervical andthoracic spine, and they are mostly dura based. PET/CT ishelpful to discover whole-body lesions and can providesome evidence for the diagnosis. Generally, pathologicalexaminations are an exclusive way to make the diagnosis.The rate of recurrence and new occurrence in other verte-brae was as high as 20%; thus, total lesion resection is per-ceived as the ideal surgical option. Radiotherapy, steroids,and/or chemotherapy are utilized individually or in caseswith residual lesions. Some cases can remit spontaneously,so a wait and watch strategy is recommended in patientswithout symptoms and signs of spinal instability andneurological impairment.

Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s13643-021-01581-0.

Additional file 1: Supplementary Table 1. Detailed information of theincluded cases with spinal Rosai-Dorfman disease in this study.

AbbreviationsRDD: Rosai-Dorfman disease; CT: Computed tomography; PET/CT: Positronemission tomography/computed tomography; FDG: Fluorodeoxyglucose;ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein; MRI: Magneticresonance imaging; TR: Total resection

AcknowledgementsWe thank LetPub (www.letpub.com) for its linguistic assistance during thepreparation of this manuscript.

Authors’ contributionsConception, P.H., F.W., X.L., and Z.L.; methodology, P.H., F.W., X.L., and Z.L.;literature searching, P.H., and F.W.; screening and inclusion, P.H., and F.W.;data extraction and collection, P.H., and F.W., X.L., and Z.L.; statistical analysis,P.H., and F.W.; validation of data and analysis, X.L., and Z.L.; resultsinterpretation, P.H., and F.W., X.L., and Z.L.; writing, P.H., and F.W.; review ofwriting, X.L., and Z.L.; tables and figures preparation, P.H., and F.W., X.L., andZ.L.; projection administration, F.W. All authors read and approved the finalmanuscript.

FundingNone.

Availability of data and materialsWe provide all supporting data in Supplementary Table 1.

Ethics approval and consent to participateNot applicable.

Consent for publicationAll authors agree with the publication of this article.

Competing interestsWe have no potential conflicts of interest.

Received: 22 July 2020 Accepted: 2 January 2021

References1. Rosai J, Dorfman RF. Sinus histiocytosis with massive lymphadenopathy. A

newly recognized benign clinicopathological entity. Arch Pathol. 1969;87:63–70.

2. Emile JF, Abla O, Fraitag S, Horne A, Haroche J, Donadieu J, et al. Revisedclassification of histiocytoses and neoplasms of the macrophage-dendriticcell lineages. Blood. 2016;127(22):2672–81.

3. Goyal G, Young JR, Koster MJ, Tobin WO, Vassallo R, Ryu JH, et al. MayoClinic Histiocytosis Working, G. (2019). The Mayo Clinic Histiocytosis WorkingGroup Consensus Statement for the diagnosis and evaluation of adultpatients with histiocytic neoplasms: Erdheim-Chester disease, Langerhanscell histiocytosis, and Rosai-Dorfman disease. Mayo Clin Proc. 2019;94(10):2054–71.

4. Nathan B, Roomallah B, Salisbury JR, Tenant-Flowers M, Kassam S. A rarecause of lymphadenopathy - Rosai-Dorfman disease in a HIV-positiveUgandan woman. Int J STD AIDS. 2016;27(12):1123–5.

5. Yao JD, McCullough AE, Walker RC, Banks PM. Brucellosis and sinushistiocytosis with massive lymphadenopathy. Am J Med. 1989;86(1):111–4.

6. Luppi M, Barozzi P, Garber R, Maiorana A, Bonacorsi G, Artusi T, et al.Expression of human herpesvirus-6 antigens in benign and malignantlymphoproliferative diseases. Am J Pathol. 1998;153(3):815–23.

7. Richardson TE, Wachsmann M, Oliver D, Abedin Z, Ye D, Burns DK, et al.BRAF mutation leading to central nervous system rosai-dorfman disease.Ann Neurol. 2018;84(1):147–52.

8. Morgan NV, Morris MR, Cangul H, Gleeson D, Straatman-Iwanowska A,Davies N, et al. Mutations in SLC29A3, encoding an equilibrative nucleosidetransporter ENT3, cause a familial histiocytosis syndrome (Faisalabadhistiocytosis) and familial Rosai-Dorfman disease. PLoS Genet. 2010;6(2):e1000833.

9. Adeleye AO, Amir G, Fraifeld S, Shoshan Y, Umansky F, Spektor S. Diagnosisand management of Rosai-Dorfman disease involving the central nervoussystem. Neurol Res. 2010;32(6):572–8.

10. Foucar E, Rosai J, Dorfman RF, Brynes RK. The neurologic manifestations ofsinus histiocytosis with massive lymphadenopathy. Neurology. 1982;32(4):365–72.

11. Sandoval-Sus JD, Sandoval-Leon AC, Chapman JR, Velazquez-Vega J, BorjaMJ, Rosenberg S, et al. Rosai-Dorfman disease of the central nervous system:report of 6 cases and review of the literature. Medicine. 2014;93(3):165–75.

12. Purav P, Ganapathy K, Mallikarjuna VS, Annapurneswari S, Kalyanaraman S,Reginald J, et al. Rosai-Dorfman disease of the central nervous system. JClinNeurosci. 2005;12(6):656–9.

13. Baassiri W, Moussalem CK, Massaad E, Zeidan YH, Darwish H. CraniocervicalRosai-Dorfman disease involving the vertebral artery: case report andliterature review. World Neurosurg. 2020;133:69–73.

14. Qin G, Ye J, Lan S, Liang Y, Xu P, Tang X, et al. Rosai-Dorfman disease withspinal and multiple intracranial involvement: a case report and literaturereview. Br J Neurosurg. 2019:1–5.

15. Lin CK, Tsai YD. Nonresectable thoracic Rosai-Dorfman Disease: a casereport and review of the literature. World neurosurg. 2019;132:309–13.

16. Kumaran SP, Thippeswamy PB, Reddy BN, Neelakantan S, Viswamitra S. Aninstitutional review of tuberculosis spine mimics on MR imaging: cases ofmistaken identity. Neurol India. 2019;67(6):1408–18.

17. Kong Z, Wang Y, Ma W, Cheng X. FDG PET/CT Image for a Rosai-Dorfmandisease with pituitary and bone involvement in a pediatric patient. ClinNuclMed. 2019;44(11):873–5.

Hu et al. Systematic Reviews (2021) 10:31 Page 8 of 10

18. Joshi SS, Joshi S, Muzumdar G, Turel KE, Shah RM, Ammbulkar I, et al.Cranio-spinal RosaiDorfman disease: case series and literature review. Br JNeurosurg. 2019;33(2):176–83.

19. Li Y, Wang X, Gao J, Yu S, Li Z. Isolated extradural Rosai-Dorfman diseasecausing the spinal cord compression: A case report. Medicine. 2018;97(40):e12722.

20. Chhabria BA, Nampoothiri RV. A quintessential syndrome with a raremarvellingaetiology: Rosai-Dorfman disease presenting as Conus-Caudasyndrome. BMJ Case Rep. 2018;2018:bcr-2017-222398.

21. Xu H, Zhang F, Lu F, Jiang J. Spinal Rosai-Dorfman disease: case report andliterature review. Eur Spine J. 2017;26(Suppl 1):117–27.

22. Karthigeyan M, Salunke P, Gupta K, Singh A, Rajasekhar R. Frozen sectioncan “sharpen” or “sand off” the surgeon’s knife: 2 case illustrations with skullbase meningioma mimics. World Neurosurg. 2017;107:1051.e1–6.

23. Carrasco-Garcia de Leon S, Flores Barragan JM, ReleaCalatayud F, BalcazarRojas O. Extra-axial mass in the foramen magnum causing cervicalcompressive myelopathy as a complication of Rosai-Dorfman disease. JClinNeurol. 2017;13(3):312–4.

24. Tu J, Li WT, Yang C. Rosai-Dorfman disease of the subdural spine with along segment lesion: A case report and literature review. J Int Med Res.2017;45(2):875–81.

25. Tripathi R, Serajee F, Jiang H, Huq A. Novel presentation of Rosai-Dorfmanhistiocytosis with a prolonged course of cranial and peripheralneuropathies. PediatrNeurol. 2017;71:70–2.

26. Rocha-Maguey J, Felix-Torrontegui JA, Cabrera-Lopez M, Gutierrez-Castro M,Montante-Montes de Oca D. A new case of cervical intramedullary sinushistiocytosis causing paraplegia and review of the literature. SurgNeurolInt.2016;7:9.

27. Igrutinovic Z, Medovic R, Markovic S, Kostic G, Raskovic Z, Tanaskovic-Nestorovic J, et al. Rosai-Dorfman disease of vertebra: case report andliterature review. Turk J Pediatr. 2016;58(5):566–71.

28. Huang BY, Zhang H, Zong WJ, Sun YH. Rosai-Dorfman disease of rareisolated spinal involvement: report of 4 cases and literature review. WorldNeurosurg. 2016;85:367.e11–6.

29. Huang BY, Liu HL, Yu CJ. Isolated Intramedullary Spinal Rosai-DorfmanDisease: A Case Report and Literature Review. World Neurosurg. 2016;88:694.e11–5.

30. de Oliveira Lima GL, da Costa AC, de Paula Goes BH, Junior NP. Doublecompression caused by isolated spinal Rosai-Dorfman disease. Spine J. 2016;16(8):e521–2.

31. Sciacca S, Barkas K, Heptinstall L, McNamara C, Shetty R. Rosai-Dorfmandisease with spinal cord compression: a diagnostic challenge. Eur Spine J.2015;24(Suppl 4):S529–35.

32. Fu X, Jiang JH, Tian XY, Li Z. Isolated spinal Rosai-Dorfman diseasemisdiagnosed as lymphoplasmacyte-rich meningioma by intraoperativehistological examination. Brain Tumor Pathol. 2015;32(1):72–5.

33. Tian Y, Wang J, Li M, Lin S, Wang G, Wu Z, et al. Rosai-Dorfman diseaseinvolving the central nervous system: seven cases from one institute.ActaNeurochir (Wien). 2015;157(9):1565–71.

34. Mannelli L, Monti S, Love JE, Kussick SJ, McLuen A, Behnia F. Primary Rosai-Dorfman disease of the bone in a patient with history of breast cancer:appearance on 99mTc-MDP scintigraphy, CT, and X-ray. ClinNucl Med. 2015;40(3):247–9.

35. Kozak B, Talbott J, Uzelac A, Rehani B. Rosai-Dorfman Disease Isolated to theThoracic Epidural Spine. J Radiol Case Rep. 2015;9(11):6–16.

36. Wu L, Xu Y. Rosai-Dorfman disease: a rare lesion with dura tail signmimicking spinal meningioma. Spine J. 2014;14(12):3058–9.

37. Kim DY, Park JH, Shin DA, Yi S, Ha Y, Yoon DH, et al. Rosai-dorfman diseasein thoracic spine: a rare case of compression fracture. Korean J Spine. 2014;11(3):198–201.

38. El Molla M, Mahasneh T, Holmes SE, Al-Khawaja D. Rare presentation ofRosai-Dorfman disease mimicking a cervical intramedullary spinal cordtumor. World Neurosurg. 2014;81(2):442.e7–9.

39. Parmar V, Seward C, Huho A, Qian J, Gandhi R, Pilitsis JG. Rosai-Dorfmandisease presenting as cervical radiculopathy. ClinNeurolNeurosurg. 2013;115(6):808–10.

40. Zhu F, Zhang JT, Xing XW, Wang DJ, Zhu RY, Zhang Q, et al. Rosai-Dorfman disease: a retrospective analysis of 13 cases. Am J Med Sci.2013;345(3):200–10.

41. Yao K, Li TF, Zhu MW, Duan ZJ, Hu ZL, Bian Y, et al. An intramedullarycervical cord lesion in a 12-year-old girl. Neuropathology. 2013;33(5):582–5.

42. Roy C, Saha A, Roy S, Ghosh A. ExtranodalRosai-Dorfman disease presentingas spinal extradural lesion: a case report with a review of the literature. JCancer Res Ther. 2012;8(4):647–9.

43. Chen CW, Kachramanoglou C, Revesz T, Choi D. Rosai-Dorfman diseasepresenting as a thoracic intradural extramedullary spinal tumor but withoutextraspinal manifestations. ActaNeurochir (Wien). 2012;154(2):367–8.

44. Antuna Ramos A, Alvarez Vega MA, Alles JV, Antuna Garcia MJ, Meilan MA.Multiple involvement of the central nervous system in Rosai-Dorfmandisease. PediatrNeurol. 2012;46(1):54–6.

45. Warrier R, Chauhan A, Jewan Y, Bansal S, Craver R. Rosai-Dorfmandisease with central nervous system involvement. ClinAdvHematolOncol.2012;10(3):196–8.

46. Zhu H, Qiu LH, Dou YF, Wu JS, Zhong P, Jiang CC, et al. Imagingcharacteristics of Rosai-Dorfman disease in the central nervous system. Eur JRadiol. 2012;81(6):1265–72.

47. Rittner RE, Baumann U, Laenger F, Hartung D, Rosenthal H, Hueper K.Whole-body diffusion-weighted MRI in a case of Rosai-Dorfman diseasewith exclusive multifocal skeletal involvement. Skelet Radiol. 2012;41(6):709-13.nguez-Baez JJ.

48. Raslan OA, Schellingerhout D, Fuller GN, Ketonen LM. Rosai-Dorfmandisease in neuroradiology: imaging findings in a series of 10 patients. AJRAm J Roentgenol. 2011;196(2):W187–93.

49. Ambekar S, Somanna S, Bhat DI, Ranjan M. Isolated cranio-spinalinvolvement of Rosai-Dorfman disease: case report. Br J Neurosurg. 2011;25(2):297–9.

50. Maiti TK, Gangadharan J, Mahadevan A, Arivazhagan A, Chandramouli BA,Shankar SK. Rosai-Dorfman disease presenting as cervical extradural lesion: acase report with review of literature. Neurol India. 2011;59(3):438–42.

51. Wang Y, Gao X, Tang W, Jiang C. Rosai-Dorfman disease isolated to thecentral nervous system: a report of six cases. Neuropathology. 2010;30(2):154–8.

52. Abou-Zeid AH, Herwadkar A, du Plessis D, Gnanalingham KK. Isolatedextradural Rosai-Dorfman disease of the thoracic spine: a rare cause ofspinal cord compression: case report. Neurosurgery. 2010;67(2):E514–5.

53. Deng XW, Xie XB, Ye YQ, et al. Intracranial Rosai–Dorfman disease: casereport and literature review. European J Radiol Extra. 2010;76:e75–8.

54. Konca C, Ozkurt ZN, Deger M, Aki Z, Yagci M. Extranodal multifocal Rosai-Dorfman disease: response to 2-chlorodeoxyadenosine treatment. Int JHematol. 2009;89(1):58–62.

55. Raslan O, Ketonen LM, Fuller GN, Schellingerhout D. Intracranial Rosai-Dorfman disease with relapsing spinal lesions. J ClinOncol. 2008;26(18):3087–9.

56. Gupta P, Babyn P. Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): a clinicoradiological profile of three cases including twowith skeletal disease. PediatrRadiol. 2008;38(7):721–8; quiz 821-2.

57. Jing X, McHugh JB, Pu RT. Fine-needle aspiration cytology of Rosai-Dorfmandisease of bone. DiagnCytopathol. 2008;36(7):516–8.

58. Dran G, Rasendrarijao D, Vandenbos F, Paquis P. Rosai-Dorfman diseasecausing spinal cord compression: case report. Neurosurgery. 2008;62(4):E977-8; discussion E8.

59. Huang YC, Tan HY, Jung SM, Chuang WY, Chuang CC, Hsu PW, et al. Spinalepidural Rosai-Dorfman disease preceding by relapsing uveitis: a case reportwith literature review. Spinal Cord. 2007;45(9):641–4.

60. Seyednejad F, Tubbs RS, Shoja MM, Daghigi MH, Oakes WJ. Presumedrecurrence of intracranial Rosai-Dorfman disease as a cervical spine tumor.ActaNeurochir (Wien). 2007;149(4):425–7.

61. Hargett C, Bassett T. Atypical presentation of sinus histiocytosis withmassive lymphadenopathy as an epidural spinal cord tumor: a casepresentation and literature review. J Spinal Disord Tech. 2005;18(2):193–6.

62. Tubbs RS, Kelly DR, Mroczek-Musulman EC, Hammers YA, Berkow RL,Oakes WJ, et al. Spinal cord compression as a result of Rosai-Dorfmandisease of the upper cervical spine in a child. Childs NervSyst. 2005;21(11):951–4.

63. Chen KT. Crush cytology of Rosai-Dorfman disease of the central nervoussystem. A report of 2 cases. ActaCytol. 2003;47(6):1111–5.

64. Yip CC, Cheng CL, Poh WT, Choo CT. Orbital, adnexal, and unusual systemicinvolvement in Rosai-Dorfman disease. Ophthalmic PlastReconstrSurg. 2002;18(3):223–7.

65. Wu M, Anderson AE, Kahn LB. A report of intracranial Rosai-Dorfman diseasewith literature review. Ann DiagnPathol. 2001;5(2):96–102.

Hu et al. Systematic Reviews (2021) 10:31 Page 9 of 10

66. Andriko JA, Morrison A, Colegial CH, Davis BJ, Jones RV. Rosai-Dorfmandisease isolated to the central nervous system: a report of 11 cases. ModPathol. 2001;14(3):172–8.

67. Hollowell JP, Wolfla CE, Shah NC, Mark LP, Whittaker MH. Rosai-Dormandisease causing cervical myelopathy. Spine (Phila Pa 1976). 2000;25(11):1453–6.

68. Kelly WF, Bradey N, Scoones D. Rosai-Dorfman disease presenting as apituitary tumour. ClinEndocrinol (Oxf). 1999;50(1):133–7.

69. Katz DS, Poe LB, Corona RJ Jr. Sinus histiocytosis with massivelymphadenopathy: a case of simultaneous upper respiratory tract and CNSdisease without lymphadenopathy. AJNR Am J Neuroradiol. 1993;14(1):219–22.

70. Unni KK. Case report 457. Sinus histiocytosis with massive lymphadenopathy(Rosai-Dorfman disease) presenting as lesion in the sacrum. Skelet Radiol.1988;17(2):129–32.

71. Haas RJ, Helmig MS, Prechtel K. Sinus histiocytosis with massivelymphadenopathy and paraparesis: remission with chemotherapy. A casereport. Cancer. 1978;42(1):77–80.

72. Kessler E, Srulijes C, Toledo E, Shalit M. Sinus histiocytosis with massivelymphadenopathy and spinal epidural involvement: a case report andreview of the literature. Cancer. 1976;38(4):1614–8.

73. Pulsoni A, Anghel G, Falcucci P, Matera R, et al. Treatment of sinushistiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): reportof a case and literature review. Am J Hematol. 2002;69(1):67–71.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Hu et al. Systematic Reviews (2021) 10:31 Page 10 of 10